Proc. Nati. Acad. Sci. USAVol. 87, pp. 3440-3444, May 1990Immunology

Human interleukin 2 receptor fl-chain gene: Chromosomallocalization and identification of 5' regulatory sequences

(promoter/Z-DNA)

JAMES R. GNARRA*, HIROKI OTANI*, MARY GE WANGt, 0. WESLEY MCBRIDEt, MICHAEL SHARON*,AND WARREN J. LEONARD**Cell Biology and Metabolism Branch, National Institute of Child Health and Human Development, and tLaboratory of Biochemistry, Division of CancerBiology and Diagnosis, National Cancer Institute, National Institutes of Health, Bethesda, MD 20892

Communicated by Ray D. Owen, February 2, 1990 (received for review December 19, 1989)

ABSTRACT Interleukin 2 (IL-2) binds to and stimulatesactivated T cells through high-affinity IL-2 receptors (IL-2Rs).Such receptors represent a complex consisting of at least twoproteins, the 55-kDa IL-2Ra chain and the 70-kDa IL-2Rchain. The low-affinity, IL-2Ra chain cannot by itself trans-duce a mitogenic signal, whereas IL-2 stimulates resting lym-phocytes through the intermediate-affinity, IL-2Rf receptor.We report here identification of the genomic locus for IL-2R,3.The exons are contained on four EcoRI fragments of 1.1, 9.2,7.2, and 13.7 kilobases. The 1.1-kilobase EcoRI fragment liesat the 5'-most end of the genomic locus and contains promotersequences. The promoter contains no TATA box-like elementsbut does contain the d(GT). class of middle repetitive elements,which may play an interesting regulatory role. The IL-2Rf3gene is localized to chromosome 22qll.2-ql2, a region that isthe locus for several lymphoid neoplasias.

T lymphocytes play essential regulatory and effector roles inthe immune response. A specific immune response is initiatedthrough the binding of cell surface antigen receptors topeptide antigens presented in association with moleculesencoded by the major histocompatibility complex. This bind-ing initiates intracellular events, including the immediaterelease of intracellular second messengers (1), the inductionof synthesis and secretion of the lymphokine, interleukin 2(IL-2) (2), and the de novo expression of surface high- andlow-affinity IL-2 receptors (IL-2Rs) (2, 3). IL-2 binding to thehigh-affinity IL-2R stimulates the events necessary for cel-lular proliferation. Expression of IL-2 and high-affinity IL-2Ris transient and requires continued antigen receptor stimu-lation for a sustained immune response (2, 3). In addition toacting as a T-cell growth factor, IL-2 can stimulate activatedB cells (4, 5), activated monocytes (6), natural killer cells (7),and large granular lymphocytes (8).Three types of IL-2Rs have been identified. High-affinity

(Kd = 10-11 M) and low-affinity (Kd = 10-8 M) receptors areexpressed on the surface of activated T cells and humanT-lymphotrophic virus type I-transformed T-cell lines (3).The intermediate-affinity (Kd = 10-9 M) receptor is ex-pressed at low levels on resting T cells and at somewhathigher levels on large granular lymphocytes (9, 10). Thelow-affinity IL-2R is composed of the IL-2Ra chain (Tacantigen, p55) and has a 13-amino acid cytoplasmic tail,presumably too short to mediate an enzymatic activity (11).The high-affinity IL-2R is composed of IL-2Ra and IL-2RB(p70) (12-14). Coexpression of these two chains yields high-affinity IL-2 binding, IL-2 internalization, and cellular pro-liferation. The f3 chain itself appears to constitute the inter-mediate-affinity IL-2R (12-14) and to be responsible for

mediating high-dose IL-2 stimulation of resting lymphocytesand large granular lymphocytes, which generate lymphokine-activated killer cell activity (10). The deduced IL-2R/3 proteinsequence (15) indicates a cytoplasmic domain large enough(286 amino acids) to potentially mediate signal transductioneither by encoding an enzymatic activity or by coupling toother proteins. Therefore, the IL-2R system appears uniqueamong multisubunit receptors in that two chains bind thesame ligand independently, but in concert form an evenhigher-affinity receptor.

MATERIALS AND METHODSScreening of cDNA and Human Genomic Libraries. A

cDNA library was prepared in A Zap II (Stratagene) fromtwice-selected, poly(A)-positive mRNA isolated from YTcells. The human genomic DNA library used was made in thelaboratory of T. Maniatis (Harvard University) and wasprepared by partial Mbo I plus Hae III digestion of fetal liverDNA and cloning into the EcoRI site of Charon 4A.

Oligonucleotides were synthesized corresponding to bases38-131 (5' probe) and 1788-1881 (3' probe) of the IL-2RSsequence (15). Probes were labeled by random priming.Duplicate filters with a total of 200,000 phage from the cDNAlibrary were differentially hybridized with the 5' and 3'probes, and three cDNA clones were selected based onhybridization to both probes. These clones were transducedinto Bluescript plasmid vectors using the helper phage, R408(Stratagene). Recombinant phage (106) were screened fromthe genomic DNA library through hybridization of duplicatefilters with either the 5' or 3' probe. Hybridizing EcoRI frag-ments were subcloned into the pBluescriptll KS(-) plasmidvector (Stratagene).Southern Blot Analysis. Total human genomic DNA from

peripheral blood cells was digested to completion withEcoRI, BamHI, Pst I, EcoRV, Taq I, Sst I, Bgl II, Msp I,HindIII, and Pvu II. Hybridization with radiolabeled probeswas for 16-20 hr at 42°C in 50%o formamide, 5x Denhardt'ssolution (lx = 0.02% Ficoll/0.02% polyvinylpyrrolidone/0.02% bovine serum albumin), 5x SSC (lx = 0.15 MNaCl/0.015 M sodium citrate, pH 7.0), and 0.1 mg of dena-tured salmon sperm DNA per ml. Membranes were washedtwice in 2x SSC/0.1% SDS at 22°C and then twice in 0.2xSSC/0.1% SDS at 500C.DNA Sequence Analysis. The 1.1-kilobase (kb) EcoRI pro-

moter fragment was subcloned into pBluescriptIl KS(-) (des-ignated pBS26) and into M13mpl8 and M13mpl9. Addition-ally, the 887-base-pair (bp) EcoRI-Bal I promoter fragmentwas subcloned into EcoRI- plus Sma I-digested M13 DNA.The plasmid pBS26 was also digested with Alu I, Hae III, orSau3A and fragments were subcloned into Sma I- or BamHI-digested M13. M13 subclones were identified by plaque hy-

Abbreviations: IL-2, interleukin 2; IL-2R, IL-2 receptor; CAT,chloramphenicol acetyltransferase.

3440

The publication costs of this article were defrayed in part by page chargepayment. This article must therefore be hereby marked "advertisement"in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Proc. Natl. Acad. Sci. USA 87 (1990) 3441

-26 1 1965 2361IL-2Rfl cDNA 5' CTGGGC'TTAGTCITJCGGCCCCAGCAGCCAGAGCTCAGCAGGGCCCTGGAGAGATGGCCA .....TCCAAACCTC .....CGAATCCTCC 3'

Published IL-2M sequence

IL-2R132 cDNA

5'GCAGCCAGAGCTCAGCAGGGCCCTGGAGAGATGGCCA .....TCCAAACCTC.......CGAATCCTCC.....3'21

5' CCCTGGAGAGATGGCCA .....TCCAAACCTC 3'

FIG. 1. Comparison of 5' and 3' endpoints of IL-2R,81 and IL-2R132 cDNA clones with the published IL-2R,8 cDNA sequence (15).

bridization with the radiolabeled 1.1-kb EcoRI fragment andused for dideoxy DNA sequencing. The sequence: was pre-pared for publication using the DNA:DRAW program (16) on theNational Institutes of Health Dec10 mainframe computer.DNA homology searches of GenBank version 60 were per-formed using the University of Wisconsin Genetics ComputerGroup package version 5 on a VAX computer located at theNational Cancer Institute, Advanced Scientific ComputingLaboratory, Frederick Cancer Research Facility.

Transfections and Chloramphenicol Acetyltransferase (CAT)Assays. The 896-bp Sma I-Bal I fragment was excised frompBS26 (using the Sma I site contained within the vectorpolylinker), HindIII linkers were ligated, and the fragment wassubcloned in both orientations into the HindIII site of Jym-CAT-0, a vector containing the gene encoding CAT but notpromoter or enhancer sequences (17). Transfections into YTor Jurkat cell lines with 5 ,ug of plasmid DNA were performedusing DEAE-dextran as described (17). Cells were harvested40-48 hr after transfection, and overnight CAT assays wereperformed with 0.2 ACi of ['4C]chloramphenicol (1 Ci = 37GBq) and acetyl CoA (18 mg/ml) (17, 18).Chromosomal Localization. DNA samples from 90 inde-

pendent human-mouse or human-hamster somatic cell hy-brids and subclones (19-21) were digested with EcoRP andthe fragments were resolved on 0.7% agarose gels. Southernblots were hybridized with the radiolabeled IL-2R131 cDNAinsert and washed as above except that high-stringencywashes were performed in 0.1 x SSC/0.2% SDS at 55°C. Thepresence of the hybridizing human sequences in the DNAsamples was correlated with the specific human chromo-somes retained in each of the somatic cell hybrids.In situ hybridization experiments were performed using

peripheral blood lymphocytes from a normal male (46; XY)that were cultured for 72 hr at 37°C in RPMI-1640 mediumsupplemented with 15% fetal bovine serum, phytohemagglu-tinin (0.5 ,g/ml), and antibiotics and then synchronized withBrdUrd (100 ,g/ml) for 16 hr. Cells were washed and resus-pended in fresh medium containing thymidine (2.5 ,ug/ml) andincubated for an additional 5.5 hr (22) with colcemid (0.05,g/ml) present during the final 20 min. After centrifugation,swelling, and fixation, air-dried metaphase spreads were pre-pared (23). ChromosomalDNA was treated with RNaseA (100,g/ml) for 1 hr at 37°C and denatured for 3 min in 0.07 MNaOH in 64% ethanol (24, 25). Radiolabeled probe (specificactivity = 3.2 x 107 cpm/,g) was prepared by nick-translationof IL-2RB1 plasmid DNA with [3H]TTP and [3H]dCTP andhybridized for 20 hr at 42°C. Slides were washed in 50%oformamide/2x SSC (pH 7.0) for 10 min at 42°C and in 2x SSCat 42°C and then coated with a 50%o solution ofNTB2 nucleartrack emulsion (Kodak). The slides were stored desiccated at4°C for 9 days and then developed, stained (0.25% Wrightstain), and photographed. The slides were then destained, andchromosomal banding was obtained by staining with 33258Hoechst (150 ,ug/ml) for 30 min and exposure to UV illumi-nation for 30 min after rinsing. The slides were again stainedwith Wright stain and the same metaphase spreads wererephotographed (22). Only metaphase spreads containing agrain on chromosome 21, 22, or Y were analyzed.

RESULTS AND DISCUSSIONAlthough the gene encoding IL-2Ra has been characterized(11) and its regulation has been studied (17), it is clear thatIL-2 responsiveness is mediated through the IL-2R/3 chain.We report here the identification ofthe IL-2RB genomic locusin order to begin to understand the regulation of IL-2Rf3 geneexpression.

Isolation of IL-2Rfi cDNA Clones. A YT cell-derived cDNAlibrary was screened based on individual plaques hybridizingto 5' and 3' IL-2R,3 probes (see Materials and Methods) inorder to select only full-length cDNA clones. Three indepen-dent clones were identified of 2.4 (IL-2R/31), 2.0 (IL-2R132)(Fig. 1), and 1.8 kb (data not shown). Their apparent identityto the published cDNA sequence was confirmed by restric-tion enzyme mapping and partial sequence analysis (Fig. 1).The longest cDNA clone extended 26 nucleotides further 5'than any IL-2Rp8 cDNA described (15), corresponding to aminor transcription initiation site (see below).Southern Blotting of Total Human Genomic DNA. Southern

blots prepared from DNA isolated from four normal humandonors were hybridized with the IL-2RP1 cDNA insert. AfterEcoRI digestion and hybridization, four bands of 13.7, 9.2,7.2, and 1.1 kb were identified (Fig. 2A). Due to the apparenthomogeneity of the four independent DNA samples and thelack of complexity of the various enzyme digestions (repre-

A B

kbp

23 -

9.4-

6.5 -_p - t

4.3-

W*2.3 -2.0-

"n'zvnie:Ll'rolbe L

C1.1-

Efco RI I Bam HIl

11-2RM1 cDNA

9.2 7.2

Eco RI I Bam HI

1.1 kb 5' fragment

+

=113.7 kbp

FIG. 2. Southern blot hybridization ofhuman total genomic DNA.(A) DNA was isolated from peripheral blood cells from four normaldonors (lanes 1-4), digested with EcoRI or BamHI, and electropho-resed on a 1% agarose gel; the DNA was transferred to nitrocelluloseand hybridized with the radiolabeled IL-2R81 cDNA clone. (B)Reprobing of the blot of A with the radiolabeled 1.1-kb upstreamEcoRI fragment shows uniform hybridization, suggesting it containsrepetitive elements. (C) Genomic organization of the IL-2RB locusshowing orientation of the four hybridizing EcoRI fragments.

MThe sequence reported in this paper has been deposited in theGenBank data base (accession no. M32979).

Immunology: Gnarra et al.

Proc. Natl. Acad. Sci. USA 87 (1990)

sentative EcoRI and BamHI digestions shown in Fig. 2A anddata not shown), these data were most consistent withIL-2Rf3 being encoded by a single or low copy number gene.

Isolation of Genomic Clones. The same oligonucleotideprobes described above were used to screen a human ge-nomic DNA A library. In this case clones were picked basedon hybridization to either the 5' or 3' probe. Five independent5' clones and two independent 3' clones appeared uniquebased on EcoRI digestion patterns. Southern blotting ofEcoRI digests from the various clones and hybridization withthe IL-2R/31 cDNA insert identified all four EcoRI fragmentsthat were detected in the human total genomic DNA South-ern analysis (see Fig. 2A). Therefore, the entire gene waslikely represented among the phage clones. The four hybrid-izing fragments were independently subcloned from the Aphage into the EcoRI site of pBluescript KS(-). Southernhybridization and partial sequence analyses identified theorder of the four EcoRI fragments from 5' to 3' (Fig. 2C). Itis possible that additional internal EcoRI fragments exist butdo not contain coding sequences.Chromosomal Localization. A somatic cell hybrid mapping

strategy was used to localize the human IL-2Rj3 gene. South-ern analysis ofDNA samples isolated from 90 human-rodentcell hybrids showed three hybridizing human bands (7.2, 9.2,and 13.7 kb) in EcoRI digests (Fig. 3), and all three bandssegregated concordantly (i.e., either all present or all absentin any given somatic cell hybrid line), consistent with theinterpretation that IL-2R,8 is a single copy gene. These bandswere well resolved from cross-hybridizing bands in Chinesehamster DNA (2.8, 5.9, 8, 11, and 15 kb) and mouse DNA(3.1, 5.6, and 15.5 kb). Analysis of the entire panel of hybrids(Table 1) permitted unambiguous assignment of the gene tohuman chromosome 22. The gene segregated discordantly(.19%) with all other human chromosomes. This chromo-somal assignment was strongly supported by the presence ofthe human IL-2Rf3 gene in a human-hamster hybrid retainingonly human chromosomes 22, X, and 6p (not shown). Re-gional localization was obtained by examination of twoindependent human-hamster hybrids containing spontane-ous breaks involving chromosome 22, both breaks occurringbetween immunoglobulin A light chain (IGL, qil) and plate-let-derived growth factor B chain (q12.1). The IL-2R,8 genewas absent in the hybrid that retained IGL but no other 22qmarker (Fig. 3, lane 26) and was present in the hybrid thatretained all 22q markers except IGL (data not shown). Theseresults permitted the regional assignment of IL-2Rf3 to 22q11-qter. In contrast, IL-2 and IL-2Ra are located on humanchromosomes 4 (26) and 10 (27), respectively.The IL-2Rp gene was localized to 22qll.2-ql2 by in situ

hybridization of human metaphase chromosome spreads.

3 5 7 9 1 1 13 15 17 19 21 23 25 C

k 9.)

13.79.27.25.9

FIG. 3. Southern hybridization of representative human-ham-ster somatic cell hybrid DNA EcoRI digests with the IL-2R,81 cDNAprobe. A different hybrid cell DNA was loaded in each numberedlane, and Chinese hamster (C) and human placental (H) DNAs arealso shown. The sizes of the hybridizing human sequences (7.2, 9.2,and 13.7 kb) and a homologous hamster sequence (5.9 kb) are shown.Weakly hybridizing 8-, 11-, and 15-kb hamster fragments are poorlyvisualized. The presence of the hybridizing human sequences isindicated (+) above the lanes.

Table 1. Segregation of IL-2R,8 gene with humanchromosome 22

Human Gene/chromosomechromosome +/+ +/- -/+ -/- % discordancy

1 15 11 14 50 282 12 14 11 53 283 17 9 16 48 284 20 6 35 29 465 16 10 7 57 196 21 5 26 38 347 13 13 26 38 438 16 10 18 46 319 18 8 12 52 2210 11 15 7 57 2411 15 11 10 54 2312 17 9 20 44 3213 9 17 23 41 4414 11 15 29 35 4915 16 10 28 36 4216 12 14 23 41 4117 17 9 36 28 5018 17 9 31 33 4419 15 11 12 52 2620 19 7 20 44 3021 20 6 41 23 5222 26 0 0 64 0X 16 10 30 34 44

The human IL-2R,8 gene was detected as 7.2-, 9.2-, and 13.7-kbfragments in EcoRI-digested human-rodent somatic cell hybrids (seeFig. 3). Detection of the gene was correlated with the presence orabsence of each human chromosome in the group of somatic cellhybrids. Discordancy represents presence of the gene in the absenceof the chromosome (+/-) or absence of the gene despite presenceof the chromosome (-/+), and the sum of these numbers times thetotal number of hybrids examined (x 100) represents % discordancy.The human-hamster hybrids consisted of 28 primary hybrids and 12subclones (17 positive of 40), and the human-mouse hybrids repre-sented 16 primary clones and 34 subclones (9 positive of 50).

Given the data from somatic cell hybrids that the IL-2Rp genemapped to chromosome 22, only metaphase spreads contain-ing hybridization to small chromosomes (i.e., 21, 22, and Y)were examined. Fifty-eight percent (19/33 total) of the grainson chromosome 22 were localized to this region (Fig. 4). Incontrast, there was a random distribution of grains on chro-mosomes 21 and Y representing nonspecific background. Anadditional 71 grains were randomly distributed on the other20 autosomes (data not shown).

Several neoplasias of hematopoietic origin have been as-sociated with nonrandom aberrations involving human chro-mosome 22 (28). Based upon its chromosomal location andfunction, IL-2Rf3 must be considered as a potential candidategene in any of these neoplasias, and genetic linkage analysiswould facilitate this evaluation. For this purpose, DNAsamples from 10 unrelated individuals were evaluated for thepresence of restriction fragment length polymorphisms(RFLPs) identified by the IL-2R4 probe. No RFLPs weredetected after DNA digestion with EcoRI, HindIII, BamHI,Msp I, Taq I, EcoRV, Pst I, Sac I, Bgl II, Xba I, and Pvu II.In Kpn I digests, a 5.2-kb band was detected in 18 of 38individuals tested and invariant bands of 12 kb and 31 kb werealso present (data not shown). Presumably one allele was notresolved from one of the constant bands in Kpn I digestionsof DNA from individuals not containing the 5.2-kb band.Moreover, it is highly probable that the d(GT)26 tract in the5' flank of the IL-2Rf3 gene (see below) is polymorphic (29,30). Amplification of this tract using the polymerase chainreaction with flanking oligonucleotide primers could makethis locus highly informative (29, 30).

3442 Immunology: Gnarra et al.

* - V, ok

.01

.440.7w

0,*

.1...;

Proc. Natl. Acad. Sci. USA 87 (1990) 3443

A

t-I.-0

'.r

il

t _

*4* I

GAATTCATGG AAATGGGAAG GGCAGTGATG GAGATGGGAA GGGCAGTGGG

AP-2GGTTGGAffG GG TTG TTGTTGGGTA CGAAAACAGA ATTAGATCGA

ATGAATAAGA TCTAGTATTT GATAGCATAA CAGGGTGACT TTAGTCAACA

ATAATTTATT GTACATTTAA AAATAACTAA AAGAGTATAC TTGGATTTTA

ACACAAAGAA AGGATAAATA CTTGAGGTGA TGGATACCCA TTACCTGATG

TGATTATTAT ACATTGTATG CCTGTATCAA AATAGCTCAT GTGCCTCATG

AATATAGACA CCTACCACAT GCCCACAAAA TTAAAAACTA AAAAAAACAG

TCATCTCTGA ATGCTAAACG GAGTAAGGGG CTTCCTGGAA GGCTGGGTGA

AATGGGAGTC TCGGAAAGAT

CGCTGGGGTC ACCTAGAGGG

GGTATGTGTG TGACCGGATG

or

*s * ~

Ob ''

1*/

X4o4000_7'

C

Soo~ ~ ~ ~ ~ ~

21 22 Y

FIG. 4. In situ hybridization. (A) Representative metaphasespread containing a grain on chromosome 22q11 (long arrow) and abackground grain on chromosome 2 (short arrow). (B) The meta-phase spread of A after chromosome banding. (C) Distribution ofgrains on chromosomes 21, 22, and Y in 47 metaphase spreads.

Identification of the IL-2R.8 Gene Promoter. Since the1.1-kb EcoRI fragment contained sequences upstream of the5' end of the cDNA, it was hypothesized to contain promotersequences. The fragment was sequenced (Fig. 5) and foundto contain 124 bp corresponding to the 5' end of the IL-2RP1cDNA and 143 bp apparently corresponding to an intron. Acanonical splice donor site is present at the indicated exon-intron boundary. Interestingly, the ATG translation start site(15) is not contained within the first exon. Transcriptioninitiation sites were mapped by S1 nuclease digestion andprimer extension. S1 digestion showed that the major tran-scription initiation site mapped to the end of the cDNA cloneidentified by Hatakeyama et al. (15) (designated + 1 in Fig. 5;data not shown) and a single primer extension experimentshowed another, minor transcription initiation site that cor-responded to the 5' end of IL-2R/31 cDNA clone (denoted atposition -26 in Fig. 1; data not shown). Examination ofthe upstream sequences showed no TATA box or CAATbox sequences. Two AP-2 sites (consensus sequence of

TCCccA CG, where N = unspecified nucleotide) (32) at

-796 to -787 and -298 to -289 and one SP-1 site (CCGCCC;ref. 33) at -163 to -158 (all boxed in Fig. 5) were identified.A second canonical SP-1 site was identified downstream ofthe major transcription initiation start site (+79 to +84).

GGTGTGTTGC AGGCTGGGAG GAGGGTGAGA

ACCTGCTTGT GTGAAGCCTA CGTATTAGTG

GAGGCGTCAG AGGTGTTGGG TAGCCTGTGT

AP-2GAGTTIGGCGT GGG TGATG TAGGAGGGGA

CCTTGGCTCC TGTGTGCAGC TAGGCCCCTA

TGTGTGTGTG TGTCTCTGTC TCTGTCTGTG

CGTAGGAGGC AGATCTTTAT CTGGCCCTGG

TTTCTCATAA GCCTCGTCTC CCCGAATCTC

GAGAGGGAGG GCCTGCGTTC

TTTGACAATG TGTCTCTGTGSP-1

TGTGTGTGTG KCCGCC CAG

GTGCTTGAGG AGTTTCAGGC

CACCCCAGCC fTGCCCCTCT-26

ATCCTCTGCA CAGGAAGTGG GCTGGCTCTG GGCTTTTAGT CTTTGCGGCC

+1CCAGCAGCCA GAGCTCAGCA GGGCCCTGGA GAGATGGCCA CGGTCCCAGC

ACCGGGGAGG ACTGGAGAGC GCGCGCTGCC ACCGCCCCAT GTCTCAGCCA

Beqinning of first intronG=IGATGTCC CCCTGCCTCC TCCCGGCCCC TGTGGACAGC CAGAGGGCTG

GGAGTGAAAG TCACAGAGAA GACTTTCAGC TCTGACTCAG TTCCCCCAGC

AGTTTCTGCC TGAACTCCCA TCCCCCAACT TTGTCTTAGA ATTC

FIG. 5. Nucleotide sequence of the 1.1-kb upstream EcoRIfragment of the IL-2R8 gene. Nucleotide numbering is based on themajor transcription initiation site (+ 1). Two AP-2 sites (-796 to -787and -298 to -289) and one SP1 site at -163 to -158 are indicatedby boxes. A second SP-1 site is 3' of the major transcription initiationsite (+79 to +84). An AP-1-like sequence (CGTCA) (31) is located at-329 to -325. The d(GT)26 Z-DNA sequences are italicized, andsequences homologous to the alcohol dehydrogenase gene subunitand ay3 region 5' flanking sequences are underlined.

Although the SP-1 sites studied are generally located up-stream of transcription initiation, SP-1 can augment activityfrom downstream locations (34). An AP-1-like, 5-nucleotidesequence previously identified as important in c-fos geneexpression was also noted at -329 to -325 (CGTCA) (31).The IL-2Rf3 upstream region also contains a stretch of 52bases containing purine-pyrimidine repeats (italicized in Fig.5). A d(GT) repeat of such length represents a family ofmiddle repetitive elements present at -105 copies in thehuman genome (35). This was confirmed by probing restric-tion enzyme-digested, total genomic human DNA with the1.1-kb EcoRI fragment and seeing uniform hybridization(Fig. 2B), indicative of hybridization to repetitive elements.As expected, comparison of this sequence to those in Gen-Bank revealed similarity to many genes containing the d(GT)"stretches. Interestingly, a d(GT)26 repeat may be expected toform a Z-DNA structure, and similar sequences have beenshown to exhibit enhancer-like function (35-37). This mayhave implications for the means ofregulation ofexpression ofIL-2Rp. In addition to the d(GT)" repeat, GenBank searchesrevealed two matches: (i) an 86% homology between -755and -657 of IL-2Rj3 and sequences contained in the 3'untranslated region of the class II enzyme of human liveralcohol dehydrogenase ir subunit mRNA (38) and (ii) a 75%homology between -722 and -618 of IL-2R/3 and 5' flankingsequences of the o-y3 region that is contained within thehuman immunoglobulin heavy chain constant region genecluster (39). Although the extent of these overlapping ho-mologies is highly significant, the function of this sequence isunclear and will require further studies.

Without classical CAAT or TATA box sequences it wascritical to demonstrate that sequences within the 1.1-kb EcoRIfragment contained promoter activity. This was done bysubcloning the Sma I-Bal I fragment from pBS26 upstream of

-804

-754

-704

-654

-604

-554

-504

-454

-404

-354

-304

-254

-204

-154

-104

-54

-4

+47

+97

Immunology: Gnarra et al.

r

Proc. Natl. Acad. Sci. USA 87 (1990)

YT Jurkat

W

~

J0 + I I PSVIIL-21Z3-C\T -CAT IL-2Rr-CAT -CAT

FIG. 6. Sequences 5' of the IL-2R3 gene have promoter activity.The 896-bp Sma I-Bal I fragment was subcloned in + and -

orientations upstream of the CAT reporter gene and used in transienttransfections of YT or Jurkat cells. The vector pSV2CAT containingthe simian virus 40 promoter and enhancer (18) served as a positivecontrol for transfections, whereas the vector Jym-CAT-O (JO) servedas a negative control for vector sequences. Percent conversions ofchloramphenicol to the acetylated form for YT cell transfectantswere as follows: JO, 0.65%; IL-2R,8(+)-CAT, 3.4%; IL-2RB(-)-CAT, 0.34%; pSV2-CAT, 94%. Percent conversions for Jurkat celltransfectants were as follows: JO, 0.17%; IL-2R13(+)-CAT, 8.9o;IL-2Rp(-)-CAT, 0.61%; pSV2-CAT, 35%.

the CAT gene. The Bal I site is located in the first exon, thusretaining the transcription initiation sites but eliminating a

portion ofthe contained cDNA sequences and the exon-intronjunction. This fragment in the correct orientation, but notreverse, was able to drive expression of the CAT reporter gene(Fig. 6). The promoter fragment was active in YT cells, whichexpress IL-2RM3, but also had significant activity in Jurkatcells, which express no detectable IL-2Rj3. Thus, additionalsequences, either 5' or 3' or within the body of the gene, mayserve an important role in controlling expression.A number of genes whose products serve "housekeeping"

functions and are expressed constitutively do not containTATA sequences. This is consistent with constitutive expres-sion of IL-2Rp8 protein (10, 40) and mRNA (15) in resting Tlymphocytes. However, IL-2Rp mRNA levels do increaseafter T-cell activation (ref. 15; unpublished results), indicat-ing an inducible regulation of expression. The nucleotideconsensus sites identified (SP-1, AP-2, and the limited AP-1-like sequence) may play a role in such regulation. The levelof IL-2Rf3 regulation differs markedly, however, from thatshown for IL-2Ra gene expression. IL-2Ra is an activationantigen: expression is not detected at the mRNA or proteinlevel in resting T cells but is greatly induced after antigenstimulation (3). Therefore, the regulation of expression ofIL-2R proteins is in keeping with the apparent functions ofthe individual subunits. IL-2Ra expression is important in theacute phase of an antigen-stimulated immune response,which is transient in nature, and may modulate (increase) theaffinity of the IL-2R for its ligand, whereas IL-2RP expres-sion is necessarily constitutive since that polypeptide appearsimportant for IL-2-mediated signaling and activation withoutantigen specificity. Future studies are necessary to addressthe coordinant and discordant means of regulation of theseseparate genes.

We thank Drs. M. Napolitano, D. Roman, and M. Toledano for

helpful discussions, Dr. J. Yadoi for YT cells, and Dr. M. A.Robinson for providing genomic Southern blots. This work wassupported in part by a National Institutes of Health Intramural AIDSgrant (to W.J.L.). H.O. was the recipient ofa grant from the JapaneseMinistry of Education, Culture, and Science.1. Weiss, A., Imboden, J., Hardy, K., Manger, B., Terhorst, C. & Stobo,

J. (1986) Annu. Rev. Immunol. 4, 593-619.2. Smith, K. A. (1988) Science 240, 1169-1176.3. Greene, W. C. & Leonard, W. J. (1986) Annu. Rev. Immunol. 4, 69-96.4. Waldmann, T. A., Goldman, C. K., Robb, R. J., Depper, J. M., Leon-

ard, W. J., Sharrow, S. O., Bongiovanni, K. F., Korsmeyer, S. J. &Greene, W. C. (1984) J. Exp. Med. 160, 1450-1466.

5. Blackman, M. A., Tigges, M. A., Minie, M. E. & Koshland, M. E.(1986) Cell 47, 609-617.

6. Malkovsky, M., Loveland, B., North, M., Asherson, G. L., Gao, L.,Ward, P. & Fiers, W. (1987) Nature (London) 325, 262-265.

7. Ortaldo, J. R., Mason, A. T., Gerard, J. P., Henderson, L. E., Farrar,W., Hopkins, R. F., Herberman, R. B. & Rabin, H. (1984) J. Immunol.133, 779-783.

8. Grimm, E. A., Robb, R. J., Roth, J. A., Neckers, L. M., Lachman,L. B., Wilson, D. J. & Rosenberg, S. A. (1983) J. Exp. Med. 158,1356-1361.

9. Dukovich, M., Wano, Y., Bich-Thuy, L. T., Katz, P., Cullen, B. R.,Kehrl, J. & Greene, W. C. (1987) Nature (London) 327, 518-522.

10. Siegel, J. P., Sharon, M., Smith, P. L. & Leonard, W. J. (1987) Science238, 75-78.

11. Leonard, W. J., Depper, J. M., Crabtree, G. R., Rudikoff, S., Pum-phrey, J., Robb, R. J., Kronke, M., Svetlik, P. B., Peffer, N. J., Wald-mann, T. A. & Greene, W. C. (1984) Nature (London) 311, 626-631.

12. Sharon, M., Klausner, R. D., Cullen, B. R., Chizzonite, R. & Leonard,W. J. (1986) Science 234, 859-863.

13. Tsudo, M., Kozak, R. W., Goldman, C. K. & Waldmann, T. A. (1986)Proc. Natl. Acad. Sci. USA 83, 9694-9698.

14. Teshigawara, K., Wang, H. M., Kato, K. & Smith, K. A. (1987) J. Exp.Med. 165, 223-238.

15. Hatakeyama, M., Tsudo, M., Minamoto, S., Kono, T., Doi, T., Miyata,T., Miyasaka, M. & Taniguchi, T. (1989) Science 244, 551-556.

16. Shapiro, M. B. & Senapathy, P. (1986) Nucleic Acids Res. 14, 65-73.17. Cross, S. L., Feinberg, M. B., Wolf, J. B., Holbrook, N. J., Wong-

Staal, F. & Leonard, W. J. (1987) Cell 49, 47-56.18. Gorman, C. M., Moffat, L. F. & Howard, B. H. (1982) Mol. Cell. Biol.

2, 1044-1051.19. McBride, 0. W., Hieter, P. A., Hollis, G. F., Swan, D., Otey, M. C. &

Leder, P. (1982) J. Exp. Med. 155, 1480-1490.20. McBride, 0. W., Battey, J., Hollis, G. F., Swan, D. C., Siebenlist, U. &

Leder, P. (1982) Nucleic Acids Res. 10, 8155-8170.21. McBride, 0. W., Swan, D. C., Tronick, S. R., Gol, R., Klimanis, D.,

Moore, D. E. & Aaronson, S. A. (1983) Nucleic Acids Res. 11, 8221-8236.

22. Bhatt, B., Bums, J., Flannery, D. & McGee, J. O'D. (1988) NucleicAcids Res. 16, 3951-3961.

23. Harper, M. E. & Saunders, G. F. (1981) Chromosoma 83, 431-439.24. Singh, L., Purdom, I. F. & Jones, K. W. (1977) Chromosoma 60,

377-389.25. Landegent, J. E., Jansen in de Wel, N., van Ommen, G.-J. B., Baas, F.,

de Vijlder, J. J., van Diujn, P. & van der Ploeg, M. (1985) Nature(London) 327, 175-177

26. Siegel, L. J., Harper, M. E., Wong-Staal, F., Gallo, R. C., Nash, W. G.& O'Brien, S. J. (1984) Science 223, 175-178.

27. Leonard, W. J., Donlon, T. A., Lebo, R. V. & Greene, W. C. (1985)Science 228, 1215-1217.

28. Trent, J. M., Kaneco, Y. & Mitelman, F. (1989) Cytogenet. Cell Genet.51, 533-562.

29. Weber, J. L. & May, P. E. (1989) Am. J. Hum. Genet. 44, 388-3%.30. Litt, M. & Lutz, J. A. (1989) Am. J. Hum. Genet. 44, 397-401.31. Ray, A., Sassone-Corsi, P. & Sehgal, P. B. (1989) Mol. Cell. Biol. 9,

5537-5547.32. Imagawa, M., Chiu, R. & Karin, M. (1987) Cell 51, 251-260.33. Dynan, W. S. & Tjian, R. (1985) Nature (London) 316, 774-778.34. Courey, A. J., Holtzman, D. A., Jackson, S. P. & Tjian, R. (1989) Cell

59, 827-836.35. Rich, A., Nordheim, A. & Wang, A. H.-J. (1984) Annu. Rev. Biochem.

53, 791-846.36. Hamada, H., Seidman, M., Howard, B. H. & Gorman, C. M. (1984) Mol.

Cell. Biol. 4, 2622-2630.37. Berg, D. T., Walls, J. D., Reifel-Miller, A. E. & Grinnell, B. W. (1989)

Mol. Cell. Biol. 9, 5248-5253.38. Hoog, J.-O., von Bahr-Lindstrom, H., Heden, L.-O., Holmquist, B.,

Larsson, K., Hempel, J., Vallee, B. L. & Jomvall, H. (1987) Biochem-istry 26, 1926-1932.

39. Akahori, Y., Handa, H., Imai, K., Abe, M., Kameyama, K., Hibiya, M.,Yasui, H., Okamura, K., Naito, M., Matsuoka, H. & Kurosawa, Y.(1988) Nucleic Acids Res. 16, 9497-9511.

40. Sharon, M., Siegel, J. P., Tosato, G., Yodoi, J., Gerrard, T. L. &Leonard, W. (1988) J. Exp. Med. 167, 1265-1270.

3444 Immunology: Gnaffa et al.